Image models - APIAny.AI Docs

Image models support one or more of:

Text to image.
Image to image with reference assets.
OpenAI-compatible image generation.
OpenAI-compatible image editing.
Gemini-compatible image workflows.

Model-specific dimensions, image size support, and reference image limits are defined by the platform capability registry and reflected in the model pages. Production integrations should prefer /v1/images/generations and poll /v1/tasks/{task_id}.

Available models

Model ID	Description
`gpt-image-2`	Image generation and editing with `auto`, positive `size` targets, or ratio shorthand mapped to the nearest official popular size, plus `low` / `medium` / `high` quality controls.
`gpt-image-2.5`	Flare-compatible default alias for fast generation and editing with six quality tiers, 1K/2K/4K, references, masks, and transparent backgrounds.
`gpt-image-2.5-flare`	Explicit low-latency Flare route for everyday generation and editing.
`gpt-image-2.5-sunburst`	Precision-focused Sunburst route for complex generation, local edits, text, and multi-reference work.
`nano-banana-pro`	Nano Banana Pro high-fidelity image generation.
`nano-banana-2`	Nano Banana 2 high-quality image generation.
`nano-banana`	Nano Banana image generation for reliable production usage.
`z-image`	Fast Z-Image generation for low-latency image tasks.
`seedance-5.0`	Seedance 5.0 image generation through Volcengine Ark.

​Available models

Available models